Manus AI Stock Analysis: Build a Full Dashboard in 3 Prompts
Manus AI built a Tesla DCF analysis and undervalued stock screener autonomously. Here's the exact 3-prompt system and what it actually produced.
Manus AI independently calculated Tesla's intrinsic value at $250. The stock was trading at $236. That's a 6% gap, and the AI got there without me touching a spreadsheet.
I ran this in one session using three plain-language prompts. What came out was a full DCF valuation, technical indicators, a SWOT analysis, and a shareable interactive dashboard, plus a separate screen identifying five undervalued stocks with fair value estimates and a sortable comparison table. Here's exactly what I did and what you can replicate.
How Manus Actually Works#
Most AI tools wait for you to tell them the next step. Manus doesn't. The moment you submit a prompt, it creates its own to-do list, starts executing tasks in sequence, checks items off as it goes, and self-corrects when something breaks.
When I ran the Tesla analysis, I could watch it work in real time on the right side of the screen: it made a directory, generated a todo.md, then worked through ten steps, gathering company data, collecting financials, running sentiment analysis, building charts, cleaning them up, and assembling a final report. By the time it was done, it had completed a full DCF model with bear, base, and bull case scenarios, all without me prompting it to do any of those individual steps.
That autonomous loop is what makes this different from asking ChatGPT a financial question and getting a wall of text back. Manus is actually doing the research, not summarizing what it already knows.
If you're not on the platform yet, here's how to get access to Manus AI, they're still in closed beta, but there's a fast path to getting approved.
The Three-Prompt System#
Prompt 1: Deep-Dive Stock Analysis#
For Tesla, I used this prompt verbatim:
"I'd like a thorough analysis of Tesla stock including a summary, financial data, revenue trends, market sentiment, technical analysis, comparable asset value, investor and investment thesis."
That single prompt triggered a ten-step autonomous research process. The output included a formatted report with a DCF model showing three scenarios. The base case intrinsic value: $250. Actual price at time of writing: $236.
The AI noted that the current stock price exceeded calculated intrinsic value across multiple scenarios, which matched the bear and bull cases it generated. The base case being this close to market price wasn't something I engineered. Manus pulled the numbers, ran the model, and landed there on its own.
Prompt 2: Build and Deploy the Dashboard#
After the analysis was complete, I gave it one instruction:
"Well done. Make a dashboard."
It started building immediately. When it hit a sandbox error, I told it to deploy to a public URL instead. It did, and within a few minutes I had a live link to a full Tesla stock analysis dashboard, executive summary, investment recommendation by investor type, financial performance charts, technical analysis, SWOT breakdown, and price target scenarios. All shareable, no login required.
That's where the quote that stuck with me came from: "with just three prompts we were able to get this beautiful dashboard and a lot of information and analysis out of it."
Prompt 3: Screen for Undervalued Stocks#
In a fresh session, I ran this:
"Find three to five potentially undervalued stocks that most retail investors are overlooking. Use your knowledge of financial analysis to determine which metrics best identify truly undervalued companies. Create a visual dashboard that clearly shows why these stocks might be undervalued compared to their true worth. Explain your findings in a way that's insightful for experienced investors but still understandable for beginners."
Manus ran seven tasks: defined the evaluation metrics, screened candidates, collected financial data, analyzed using PE, PB, EV/EBITDA, and PEG ratios, then built and deployed the dashboard.
The five stocks it identified: General Motors, PulteGroup, D.R. Horton, Synchrony Financial, and United Airlines.
The published dashboard included a key metrics comparison table, sector distribution chart, individual stock graphs, a composite score for each company, current price, and an estimated fair value, all with interactive sorting built into the HTML.
Where to Be Careful#
The dashboard is impressive. The limitation is real: Manus generated an "estimated fair value" for each stock but didn't explain the methodology behind it. I noticed this. The composite score also appears without a breakdown of how it was calculated.
If you're using this for actual investment decisions, you need to audit those numbers. Ask Manus to explain its fair value calculation in a follow-up prompt, or treat the output as a starting point for deeper research rather than a final answer. The screening and visualization layer is genuinely useful. The specific price targets deserve scrutiny.
The PEG ratio averages it included in the comparison table are also a bit odd, averaging growth rates across different companies in different sectors doesn't tell you much. That's a judgment call the AI got slightly wrong, and it's worth knowing those edge cases exist.
What This Actually Means for Your Workflow#
The practical use case here isn't replacing a financial analyst. It's eliminating the part of research that takes the most time: gathering data, building models, and formatting outputs into something readable and shareable.
A workflow that used to require financial modeling skills, multiple data sources, and hours of spreadsheet work now takes three prompts and a single session. The output is good enough to brief a team, identify candidates for further research, or pressure-test a thesis you already have.
For a deeper look at how tools like this fit into a broader AI stack, the best AI tools for solopreneurs in 2026 post covers what's actually worth paying for versus what's hype.
The three prompts above are ready to use as-is. Start with a stock you already have a view on, that way you can immediately sanity-check what Manus produces against what you already know, and you'll learn faster where to trust it and where to push back.
If you found this useful, this video goes deeper on getting started with the platform:
Watch the full video on YouTube: https://youtu.be/bpwRubx_cc4
This post contains affiliate links. I only recommend tools I actually use.
Get new videos in your inbox
Weekly AI workflows. No fluff.
No spam. Unsubscribe anytime.